Global Technology Watch

CHIPS, Summer, 2004 by Ronald Neil Kostoff

Overview

Governments and industry worldwide rely on advances in science and technology (S&T) to maintain a competitive advantage. To this end, they need ready access to the results of global research to:

--Track the impact of research to help identify benefits

--Evaluate science and technology programs

--Avoid research duplication

--Identify promising research directions and opportunities

--Perform myriad oversight tasks

--Support every step of a strategic research process that makes optimal use of S&T investment resources

In addition, recent counterterrorism concerns highlight the need for ready access to information that links people, technology and organizations together to stop the threat of terrorist activities. To combat this threat, more advanced technology is required, especially in the areas of surveillance, detection and prediction.

Since science and technology are global enterprises, with expenditures approaching $1 trillion dollars annually, (depending on one's definition of S&T), no single organization or nation, can begin to research and develop the full spectrum of S&T required for a modern competitive economy or military. There must be cooperative development efforts including identifying, leveraging and exploiting external efforts--if an organization or nation is to remain competitive.

Global Technology Watch maintains awareness at all levels of global S&T through a combination of human-based overt and covert activities, and automated approaches for analyzing and tracking the myriad S&T outputs. These outputs include text (reports, papers, patents, etc.), other media, physical products and technically trained people.

This article describes how information technology can help an organization maintain awareness of global S&T efforts by extracting useful data from large volumes of structured and unstructured S&T text. It is targeted to the researcher, intelligence analyst and information technology professional.

Powerful information technology techniques, such as text mining, now exist to identify and extract relevant data from the global S&T literature. Text mining is especially useful in making sense out of disjointed and disparate data. At the Office of Naval Research, we have developed and used these techniques to substantially enhance the retrieval of useful information from global S&T databases, such as the following.

* Science Citation Index (SCI)--current and retrospective bibliographic information, author abstracts and cited references found in 5,600 of the world's leading scholarly science and technical journals covering more than 150 disciplines. The Web-based Science Citation Index Expanded, used at the Office of Naval Research, has 2,100 more journals than the CD-ROM version.

* Engineering Compendex--a compendium of more than 5,000 journals, conference proceedings, technical reports and foreign translations addressing applied research and technology development.

* MEDLINE--published by the National Library of Medicine and the National Institutes of Health, containing medical data covering basic and applied research.

* National Technical Information Service (NTIS)--the largest central resource for government-funded scientific, technical, engineering and business related information available today with more than 600,000 information products covering over 350 subject areas from over 200 federal agencies, including the Defense Technical Information Center Technical Reports. The technical reports and other DTIC databases are easily accessible on the DTIC Web site at http://www.dtic.mil/.> * Inspec--published by the IEE, is an English-language bibliographic information service providing access to the world's scientific and technical literature in physics, electrical engineering, electronics, communications, control engineering, computers, computing, information technology, manufacturing and production engineering.

* RADIUS--created by the Rand Corp., in cooperation with the National Science Foundation, contains narratives of U.S. government agency research and development programs at five hierarchical levels, ranging from 24 narratives at level 1 (reflecting overall descriptions of the research and development activities of the 24 major R&D sponsoring agencies) to 592,000 narratives at level 5 (award levels from these 24 agencies).

* U.S. Patent and Trademark Office--patent database.

The extracted data is used to identify the technology infrastructure, including authors, journals, organizations, etc., of a technical domain and the experts for innovation-enhancing technical workshops and review panels. It is also used to:

* Develop site visit strategies to assess organizations globally using bibliometrics (e.g., counts of publications, patents, citations and unpublished data) and other science and technology indicators.

* Generate technical taxonomies (classification schemes) using clustering methods.

* Provide roadmaps for tracking innumerable research impacts across time and applications areas based on text mining. This has important consequences for Web-based corporate and national security intelligence.